Prediction of breakdown voltages in N-2+SF6 gas mixtures


Tezcan S. S. , Dincer M. S. , Hiziroglu H. R.

Annual Conference on Electrical Insulation and Dielectric Phenomena, Missouri, United States Of America, 15 - 18 October 2006, pp.222-225 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/ceidp.2006.312101
  • City: Missouri
  • Country: United States Of America
  • Page Numbers: pp.222-225
  • Gazi University Affiliated: Yes

Abstract

This study proposes artificial neural networks (ANN) to predict the breakdown voltages in N-2+SF6 gas mixtures. The proposed ANN consists of one input layer, two hidden layers and one output layer, which is essentially the predicted breakdown voltage. In order to train the ANN, the experimental data available in literature for N-2+SF6 have been used. When compared with the experimental data the average relative errors on predicted breakdown voltages are found to be less than +/- 5% for training as well as for testing in all cases using the proposed ANNs. Since the average errors are less than 5%, it is recommended to use the proposed ANNs to predict the breakdown voltages.